Hello, I’m fully newbie on Hailo 8 module. I recently want to use the pretrained petrv2 models directly and want to combine with rpicam cli to get the first step results. Is there any method to write post process file to use it? Or should I additionally declare a class to achieve that?
Please take a look in this repository:
Hailo-BEV Repository.
In this repo, we run the PETRv2 model on Hailo-8 with x86 or automotive platforms. However, we haven’t tested it on an RPi. The demo utilizes data from the nuScenes dataset, not from cameras.
Please note that the PETRv2 model requires input from six cameras, trained on a specific camera arrangement, and is not compatible with a single RPi camera.
Let me know if you have further questions!
Best regards,
Ronit
Thanks for your response. I’ll go to check more detail about it.